Harmonious bilingual development: Young families’ well-being in language contact situations

Different languages: The use of dual-lingual conversations

When interlocutors speak different languages in the same conversation, that is, when they each speak a different language, but understand the other language, the conversations are dual-lingual (Saville-Troike, 1987). In public life, such dual-lingual conversations have been advocated as excellent ways for overcoming the problem that bilinguals may not be proficient enough to actually speak a particular language (e.g. ten Thije & Zeevaert, 2007). In the intimacy of family discourse, however, dual-lingual conversations may be problematic. For example, in one case the boy Orren heard both Hebrew and English from birth (Karniol, 1992). His parents, sister and visiting relatives addressed him only in English. Amongst themselves and with Orren’s sister, Orren’s parents spoke Hebrew. The parents expected Orren to speak English when they addressed him in English. At first, they did not accept it when he started to respond to them in Hebrew. As time went on, though, Orren just spoke Hebrew in response to English. Orren repeatedly begged his parents to speak Hebrew to him, but they refused. In extracts of transcripts in Karniol’s (1992) article Orren clearly expresses frustration, sadness and anger because his parents would not speak any Hebrew to him (and his parents did not like him to speak Hebrew to them). For Orren and his family, well-being was being threatened because of the language contact situation. In particular, it was the language choice in conversation that appeared to be at the root of discord. The example from Orren also shows that different people may have different perspectives on the use of a particular language in a given situation. Language choice patterns within the family, then, are far from neutral (De Houwer, 2009; Portes & Rumbaut, 2001). As Tseng and Fuligni (2000: 473) wrote on the basis of their survey of 600 adolescents from immigrant families in the United States: “Adolescents who conversed with their parents in different languages felt more emotionally distant from them and were less likely to engage in discussions with them than were youths who shared the same language with their parents.” Such emotional disengagement is possible because, as Hua (2008) has shown for families with older children, language choice patterns within the family have great symbolic meaning. They affect and are created by the dynamics of family relationships.

The very fact that there are no studies that critically discuss the use of the same language within family discourse suggests that not using dual-lingual conversations is commonly regarded as unproblematic.

Bilingual children speaking just a single language

Even when dual-lingual conversations within the family are not experienced negatively by any of the interlocutors, if children engage in them all the time they may effectively be limiting themselves to just a single language whenever they speak. This happened to Kate, who stopped speaking Dutch at the age of 4;0 (years;months) after she had moved from Belgium to the US with her family at the age of 3;8. In the US her father continued to speak Dutch to her, but Kate started to answer just in English. When at age 4;6 Kate visited Belgium again, she could no longer communicate with her monolingual Dutch-speaking grandparents, who understood no English (De Houwer, 2009). On tape recordings that her grandfather made, Kate responds in English to anything her grandfather says to her in Dutch, and there is a lot of anger that comes through her English utterances. Kate’s father felt guilty about the fact that Kate could no longer communicate with her grandparents. The family’s well-being was adversely affected because of the bilingual situation, and the prior use of dual-lingual conversations with her father where Kate spoke only English had the ultimate result that there was no longer any communication possible between Kate and her monolingual grandparents, who were unable to engage in dual-lingual conversations.

Although parents may differ in their acceptance of dual-lingual conversations, parents who initially had no problems with their child answering in a language other than the one they were addressed in may start to become worried when children are aged three or four and are speaking only a single language. As children become more competent speakers and gain more maturity, parents may start to feel that part of their identity is being lost when their children do not share their language. Parents will also begin to realize that communication between their child and relatives who speak only the other language may become impossible. Children themselves may become embarrassed in interactions with relatives whose language they cannot speak. Indeed, if children speak only a single language in spite of parental input in another language, it may be impossible for children to communicate with members of their extended family and/or family friends.

When children do not speak their parents’ language, communication between parents and children may also be adversely affected (Wong Fillmore, 2000). A case in point is that of Lauren, who heard English and Dutch from birth and grew up in the Dutch-speaking region of Flanders in Belgium (De Houwer, 2009, 2011). Lauren heard English solely from her American father and occasionally from friends he brought home. Lauren’s father worked long hours and did not spend much time with Lauren. At age three, Lauren spoke fluent Dutch and understood some English, but she said only “yes” and “no” in English. This greatly angered her father, who claimed that by not speaking his language Lauren was rejecting him. Clearly, Lauren’s early bilingualism was not a case of harmonious bilingual development.

When children and parents do not share a language, both parents and children may feel a sense of loss of identity, culture, and emotional bonding (Kouritzin, 1999; Portes & Hao, 1998; Tseng & Fuligni, 2000). Furthermore, if in bilingual communities many children speak just a single language, language shift may be the result (e.g. Fishman, 1991; Kulick, 1992). Communities may experience language shift as a tragic or at least deplorable event that does not contribute to a community’s well-being (e.g. Gafaranga, 2010; Kulick, 1992).

These examples of potentially grave and lasting negative consequences support the claim that not speaking two languages in an early bilingual setting hinders harmonious bilingual development. Actively speaking two languages appears to be the expected default, similar to actively speaking one language in a monolingual setting. Kate fulfilled this expectation before the family moved to the US (De Houwer, 1990), as do the bulk of preschoolers reported on in the bilingual acquisition literature (e.g. Qi, 2011; Serratrice, 2001; Unsworth, 2003). This literature is focused primarily on children with bilingual input from birth (De Houwer, 2009).

Also children who grew up first with one language and added a second language later on may stop speaking their first language. This was the case for Michal, who first acquired Hebrew in Israel (Kaufman & Aronoff, 1991). At age 2;6 she moved to the United States with her parents and two older siblings. At home, the family continued to speak Hebrew. Michal started attending preschool and started to learn English as a second language. By age 4;6, Michal had stopped talking Hebrew except for some mixed utterances combining Hebrew and English words. When she was addressed in Hebrew she responded in English (or with a mixed utterance). The authors (Michal’s parents) do not give any information on whether this loss of Hebrew was seen as problematic either by them, Michal’s siblings or Michal herself. We hence do not know whether there was harmonious bilingual development or not. One would have to ask the people involved in the communicative setting. There is, however, a definite possibility that because of the partial loss of Hebrew the bilingual setting was not experienced entirely positively by all family members.

Although in an immigration setting gaining second language proficiency will contribute to well-being (e.g. Liebkind & Jasinskaja-Lahti, 2000), high proficiency in both the heritage language and the second language offers adults significantly higher chances of experiencing well-being than proficiency in either of these just by themselves (Schachter, Kimbro, & Gorman, 2012). Similarly, Liu et al. (2009) found that heritage language proficiency was a protective factor against depression in foreign-born teens. Six-year-old children of immigrants were rated higher by their teachers on measures of emotional and behavioral well-being if they had good skills in both their L1 and their L2 (Collins et al., 2011). These findings support the assumption here that harmonious bilingual development is crucially linked to speaking two languages rather than just one (see also Han & Huang, 2010).

Differential proficiency in two languages

If children actually do speak two languages, they do not necessarily speak them with the same levels of fluency or proficiency. One language may be more weakly developed than the other. In fact, such uneven development appears to be quite common (De Houwer, 2009), although statistical information on this is lacking.

Parents of children who clearly speak at different levels in each of their languages may not mind, and may be pleased that their children are speaking two languages at all. On the other hand, parents may worry that children are not developing as fast in one of their languages as one might wish. In a bilingual situation, people are quick to attribute any perceived language learning problems to the bilingual setting, even though the problems may have nothing to do with it. This conclusion is unfortunately often drawn by many speech and language professionals, members of the medical community, and other professionals dealing with young children (Baker, 2011; De Houwer, 2009). Children themselves may feel embarrassed that they cannot express what they want in one of their languages, and for this reason children may start to avoid speaking one of their languages.

Many adults who grew up with two languages from early on feel inadequate because they are constantly comparing their own bilingual language use to some “monolingual” ideal. In particular, they feel dissatisfied that they have been unable to learn to write in two languages at equal levels that are on a par with what they view as expected from monolinguals, and people blame this feeling on the fact that they were raised bilingually (these observations are based on many personal communications to the author). Similarly, many parents, educators, and language and speech professionals expect bilingual children to function like monolingual children. What Wölck wrote in 1988, viz. "nearly all of the so-called Limited English Proficiency (LEP) children from ethnolinguistic minorities in U.S. public schools are still expected to reach the standards of their monolingual English-speaking schoolmates" (Wölck, 1988: 5), still appears to be generally true for the US and, mutatis mutandis, for many other countries. Harmonious bilingual development may be at risk if children do not live up to these "monolingual" expectations.

There have to my knowledge not been any systematic studies of how young children and their families experience uneven development. The few case studies of children with a clearly stronger and weaker language (see e.g. Bonnesen, 2009, for two children exposed to two languages from birth and Montanari, 2005, for three early second language learners) tend to focus on linguistic aspects only, rather than on feelings children and their families might have regarding the fact that one language is clearly less developed than the other one.

On the other hand, there is clear evidence from a child who had about equal proficiency in two languages for the notion that the lack of clearly uneven development supports harmonious bilingual development. Kate heard both English and Dutch from birth in a one person, one language situation where her American mother addressed her in English and her bilingual Flemish father in Dutch (De Houwer, 1990). Kate went to an English-speaking preschool. Her grandparents spoke just Dutch. Up until age 3;8, Kate lived in Flanders, the Dutch-speaking region of Belgium. She understood and spoke both English and Dutch without any clear differences in proficiency in either. Kate’s early bilingualism was just a given that nobody except for me, the researcher, found of any particular interest. Kate spoke Dutch with her paternal grandparents, her father, myself, and anyone else who addressed her in Dutch. She spoke English with her mother, at school, with other children and English-speaking international visitors. There was no indication that the bilingual situation caused anyone any problems, or that the family’s well-being was negatively affected by the bilingual situation. We may thus say there was harmonious bilingual development.

The following section explores possible explanations for the elements identified here as central to harmonious bilingual development when positive attitudes to early bilingualism are in place: the occurrence of conversations in which parents and children speak the same language, children speaking two languages, and the attainment of fairly equal levels of proficiency in two languages.

Conversations in a single language

If it is socially acceptable for children to answer in a language other than the one they are addressed in they may have no need to actually speak two languages. The social acceptability of particular language choices is communicated to young children by their parents’ and caretakers’ discourse strategies (Lanza, 1992, 1997). Through interaction patterns, young children are socialized into learning what is expected of them in terms of language choice.

Parents who understand and speak only one of the languages a bilingual child is exposed to will necessarily be using what Lanza (1992, 1997) has called monolingual discourse strategies. That is, parents will stick to using a single language themselves, and will show a lack of understanding when they hear anything that does not sound like the particular language they expect to hear. With a monolingual parent, then, children are forced to follow the parent’s language choice if they wish to communicate with the parent at all (which young children normally want to do).

Parents who understand both languages that a child is hearing have more and different options. They can use monolingual or more bilingual discourse strategies (Lanza, 1992, 1997). That is, through conversational reactions they can communicate to children that they are supposed to use only a single language in a particular setting (monolingual discourse strategies), or they can communicate to children that it is fine to use either language (bilingual discourse strategies). An example of a monolingual discourse strategy would be to say something like “what?” if the child asked something in the language the parent does not usually address the child in. An example of a bilingual discourse strategy would be to give a child apple juice even if she asked for it in the language the parent does not usually address the child in.

Note that parents can use two languages in speaking to a child and still use monolingual discourse strategies in each, that is, parents can socialize children into speaking language X when they are addressed in language X, and language Y when they are addressed in language Y.

Parental monolingual discourse strategies in each of two languages support children’s active use of two languages. Parental bilingual discourse strategies in either language do not support children’s active bilingual use. The use of bilingual discourse strategies with young bilingual children can quickly lead to the creation of dual-lingual conversations. If parents allow these, children may soon no longer use more than one language.

The empirical basis for the potentially pivotal role of parental discourse strategies so far consists of only a handful of case studies involving a total of not even 10 children (e.g. Juan-Garau & Pérez-Vidal, 2001; Kasuya, 1998; Lanza, 1997). Yet these studies are very promising, since some of them show that bilingual parents who initially used bilingual discourse strategies and allowed dual-lingual conversations managed to turn things around after they started using just monolingual discourse strategies. Whereas previously their children spoke just a single language, after parents changed strategies children were actively using two languages (Juan-Garau & Pérez-Vidal, 2001; Kasuya, 1998; Taeschner, 1983; see, however, Meng & Miyamoto (2012) for an example of where changed strategies appeared not to have any effect).

It must be noted that the children in these case studies were under four years of age. When children are older it may be much harder to change language use patterns and to get children to start speaking two languages. Parents may even feel it is a lost cause (see De Houwer, in press).

While language socialization is initiated by parents, children themselves play an active role as well. Three-year-old Dutch–English bilingual Susan, who heard Dutch from her mother and English from her father, started to cry when her mother inadvertently addressed her in English. This contributed to her mother making sure to address Susan only in Dutch (De Houwer, 2009). Gafaranga (2010) gives an account of older children interacting solely in French with their parents in a Kinyarwanda–French bilingual setting; these older children were managing to push through their own language preferences and parents seemed to have given up on getting children to speak Kinyarwanda. Instead, parents were using more and more French themselves.

Statistical data on the extent to which bilingual families with young children use dual-lingual conversations is lacking, as is information on the frequency of use of bilingual vs. monolingual discourse strategies. As indicated earlier, the use of dual-lingual conversations could be an important precursor for children’s lack of active bilingual use. Fortunately, empirical data on children’s active bilingual use or their lack of it do exist. These I turn to next.

Children’s active bilingual use

When children are raised with just a single language, they as a rule learn to speak the language they are exposed to. The very small proportion of children who do not are usually considered speech- or language-impaired (Leonard, 1998). In a bilingual setting, children may be speaking one language rather than two without being speech- or language-impaired.

Speaking one language rather than two in a bilingual setting is not exceptional: in a large sample of second-generation teens between 12 and 17 in the US, less than a third were fluent in both English and their heritage language (Portes & Hao, 1998). A study focusing on bilingual families with children between six and nine years of age and their siblings found that as many as a quarter of 5,320 children raised with two or more languages spoke just a single language (De Houwer, 2003). It is invariably the majority (school) language that bilingually raised children and teens speak when they limit themselves to only a single language (e.g. De Houwer, 2003; Portes & Hao, 1998; Schecter & Bayley, 1997; Yamamoto, 2001).

The implication of the high proportions of bilingually raised children and teens speaking just a single language is that many must address their parents in a language other than the one the parents speak to them. This means there is a large chance of frustrated or conflictive bilingual development – if not for the children, then for the parents. It is thus important to try and explain why in spite of bilingual input so many children speak only a single language.

There are various reasons why children may speak only a single language in spite of bilingual language input. Studies by Park et al. (2012), Portes and Hao (1998), Sirèn (1991) and Yamamoto (2001) have identified parental language choice patterns as an important factor. My own study of parental language choice patterns in relation to bilingual children’s active bilingual use has expanded on these studies and confirmed their basic findings (De Houwer, 2007). As shown in Table 1, actively bilingual children can mostly be found in families where both parents speak just a minority language at home (children in these families hear the majority language outside the home). However, most of the families in the survey have parents who employ two languages in the home.

OPEN IN VIEWER

Table 1.

Parental home language choice patterns and children’s active bilingual use.

Parent 1 speaks	Parent 2 speaks	Proportion of children who speak two languages
MinL	MinL	97%
MinL + MajL	MinL	93%
MinL + MajL	MinL + MajL	79%
MinL	MajL	73%
MinL + MajL	MajL	34%

MinL: minority/heritage language; MajL: majority/school language. (adapted from Table 7 in De Houwer, 2007).

There are different ways in which two home languages are distributed amongst the parents. Some distribution patterns are more successful than others; that is, some home language use patterns are more strongly associated with children actively speaking two languages than others. For example, if both parents speak the minority language and one of the parents in addition speaks the majority language, the chances that the children will speak two languages are greatest. Parents who employ two languages at home but who show the reverse pattern, i.e. they both speak the majority language at home but one parent in addition also speaks the minority language, stand the most chance of having children who just speak a single language.

The underlying reasons for why specific parental input patterns better correlate with children’s active bilingual use than others are not clear (De Houwer, 2007). However, the opportunities for children to speak the minority language are greater when two parents use the minority language at home, and since children are incipient language learners they need practice in order to become fluent speakers of a language (Clark, 2003). This may explain why the three parental input patterns where both parents speak the minority language at home have a higher association with children who speak two languages than the two patterns where just a single parent speaks the minority language (Table 1).

Different parental input patterns may also reflect different constellations in the absolute frequency with which children hear the minority language, that is, in how often they hear it spoken (De Houwer, 2007). This input frequency is probably higher when two parents speak the minority language than when only one parent does.

The relationship between input frequency and children’s active bilingual use has not been systematically investigated. However, the totality of anecdotal reports on individual bilingual children’s development is highly suggestive of a connection (De Houwer, 2009): the more children hear a particular language, the more chance there is that they will actually speak it. The fact that school-aged children generally speak the language they hear most frequently (viz. the majority language) but not necessarily the minority language is also suggestive of a connection. Actual conversational data for a small group of English–Inuktitut bilingual children in conversations with their parents and other adults also suggest that the more often adults speak a particular language, the more children will use that language as well. Data from Allen, Genesee, Fish, and Crago (2002) and Allen (2007) that I used for a new analysis (De Houwer, 2011) show a near perfect match between the rank orders for the children and their adult caregivers: children spoke more English with adults who spoke more English compared to others, and children spoke more Inuktitut with adults who spoke more Inuktitut compared to others.

Children speaking two languages at similarly good levels

Although systematic studies are lacking, it appears that uneven development in the two languages that bilingual children actually speak is the norm rather than an exception (De Houwer, 2009; Paradis, Genesee, & Crago, 2011). It is thus important to consider the factors that lead to uneven development.

New research is emerging that suggests input frequency as an important factor. The language that children hear most frequently is the one they will develop a larger lexicon in (Hoff et al., 2012; Nakamura & Quay, 2012; Pearson, Fernández, Lewedeg, & Oller, 1997). Blom (2010) and Hoff et al. (2012) present data showing that grammatical development in young bilinguals may also be related to the relative input frequency in both languages. Several children discussed in the literature who were reported to have about equal levels of proficiency in each language heard their two languages to more or less the same extent (De Houwer, 2009). As suggested in De Houwer (2014), though, not only relative input frequency may be of importance (which language is heard more often), but also absolute input frequency (how much is actually spoken to children in each language). Not only the input quantity but also the quality of language input may have an effect on bilingual children’s proficiency in each language (Quiroz, Snow, & Zhao, 2010).